A Systematic Empirical Analysis Of Chinese Interlanguage Based On Word Co-occurrence Network

If we concede that Chinese interlanguage exists,it is tantamount to admitting that it is an independent system,but at present,the systematic nature of this interlanguage is still a hypothesis and lacks the support of empirical research.If we want to show that Chinese interlanguage has systematic characteristics,we should not only stay at the level of a "hypothesis",but also support definite motivation.The complex network of language is an implementation of the systematic view of language,which provides a quantitative method for studying the overall characteristics of language at a systematic level.The cooccurrence relationship of words can reflect the morphological and syntactic features of a language.Modeling the Chinese interlanguage as a network of cooccurrence of words can reveal the overall characteristics of the Chinese interlanguage at a systematic level,thus making a genuine systematic analysis of the proposition of "the systematic nature of the Chinese interlanguage".This paper adopts the complex network method of econometric linguistics.Firstly,based on the HSK dynamic composition corpus of Beijing Language and Culture University,approximately 210,000 words of Chinese learners’ compositions from six countries are selected,and approximately 100,000 words of Chinese interlanguage and Chinese mother-tongue corpus are collected based on the same composition topic.Then,the corpus in the selfbuilt corpus is modeled as a corresponding word co-occurrence network.Secondly,according to the language complex network analysis index,the word co-occurrence networks of the Chinese interlanguage,Chinese mother tongue and Chinese learners of different native languages are compared and analyzed.This paper analyzes the overall characteristics of the Chinese interlanguage word co-occurrence network,and conducts an in-depth analysis of the different characteristics of the different subsystems of the Chinese interlanguage system.Thirdly,the complex network research method is further used to analyze the systematic characteristics of the Chinese interlanguage from the two dimensions of small world attribute and scale-free characteristic.Finally,the characteristics of the Chinese interlanguage system are described in detail.Through the comparative analysis of the network parameters,this paper finds that: first of all,on the whole,the co-occurrence network of Chinese interlanguage and Chinese native words has some similar characteristics,and the network density and network aggregation coefficient are very close.Secondly,there is a certain heterogeneity in the co-occurrence network of Chinese interlanguage words.Under the same premise,the vocabulary of Chinese second language learners is more abundant and diverse than that of Chinese native language learners,the process of vocabulary output is more active,and the ability of word co-occurrence is stronger.Third,both the Chinese interlanguage word co-occurrence network and the Chinese mother tongue word co-occurrence network have small-world attributes and scale-free characteristics,which strongly proves that Chinese interlanguage is a system and has systematic characteristics.Finally,it comprehensively reveals the characteristics of the Chinese interlanguage system,that is,the Chinese interlanguage vocabulary network is generally "dense" and "efficient" and the aggregation of the network is relatively low.There are differences in network compactness among Chinese learners in different countries.